Building Skill-Centric AI Products with Yogendra Miraje

AI Engineer17m 24s
0 comments · 0 votesOpen discussionClose discussion
Sign in to join the discussion

    Video summary

    Yogendra Miraje describes FactSet's move from proprietary blueprints to an open skills format. He separates prompts, which define an agent's role, from tools, which determine its connections, and skills, which describe how tasks should be completed.

    His minimal harness combines a registry, system prompt, file-reading tool and agent loop. A financial-research example demonstrates on-demand instructions and distinct HTML or PDF routing. Miraje recommends organizing skills around user intent rather than underlying datasets.

    A model upgrade broke unchanged skills because important instructions appeared late in the file. That example motivates model-specific evaluation, named maintainers, tool access boundaries and versioning. For larger libraries, he discusses retrieval and hierarchy while framing governance as practical maintenance rather than unnecessary bureaucracy.

    Original YouTube thumbnailWatch on YouTube

    Share this page

    Yogendra Miraje in a blue top beside the blue and white headline Skills Are Features on a black background. Framed in blue with WWW.ARTIFICIAL-INTELLIGENCE.VIDEO, 29 July 2026 and duration 17m 24s.

    Yogendra Miraje shows how skills can become agent-product features, supported by a small harness, clear routing descriptions and evaluations tied to the running model.